Figure 4. Effects of adding a constraint on the tutor firing rate to the simulations.

(A) Learning was slowed down by the firing rate constraint, but the accuracy of the final rendition stayed the same (inset, shown here for one of two simulated output channels). Here $\alpha =0$, $\beta =-1$, and ${\tau }_{\text{tutor}}={\tau }_{\text{tutor}}^{*}=40\mathrm{ms}$. (See online Video 3.) (B) Sequential learning occurred when the firing rate constraint was imposed on a matched tutor with a long memory scale. The plots show the evolution of the motor output for one of the two channels that were used in the simulation. Here $\alpha =24$, $\beta =23$, and ${\tau }_{\text{tutor}}={\tau }_{\text{tutor}}^{*}=1000\mathrm{ms}$. (See online Video 4.).

DOI: http://dx.doi.org/10.7554/eLife.20944.007